[Paper] Packaging Design Analysis by Predicting User Preference and Semantic Attribute

Packaging design has a pronounced effect on consumer purchase behavior and can be critical factor in marketing. Despite the importance, there are very few studies that have investigated optimal designs. In this work, order to analyze packaging designs support designing processes, we propose deep learning based method with ensemble predict user preference for design. For qualitative analysis, visualize feature maps from prediction model. Moreover, mentioned frequencies of semantic attributes questionnaires, which represent impressions users have. The experimental results suggest terms prediction, proposed model achieves correlation coefficient 0.652 between ground truth scores predicted values. As highest reaches 0.75. Also visualization successfully indicates key elements